AttributeError: module 'nbconvert.preprocessors.execute' has no attribute 'CellExecutionError'

[nilswagner]

How can I resolve the following issue?

nbsphinx/src/nbsphinx.py", line 995, in parse except nbconvert.preprocessors.execute.CellExecutionError as e: AttributeError: module 'nbconvert.preprocessors.execute' has no attribute 'CellExecutionError'

[mgeier]

Are you using a pre-release version of nbconvert?

See also https://github.com/jupyter/nbconvert/issues/1237.

[nilswagner]

nbconvert.version '6.0.0a3'

[mgeier]

Possible work-arounds:

• use nbconvert < 6
• apply this patch: https://github.com/mgeier/nbsphinx/commit/b90cfb5d2a8cabf70d5dd21d1c105e71a334dc06
• ask the nbconvert people to provide a smooth transition instead of a hard breaking change, see https://github.com/jupyter/nbconvert/issues/1237

I cannot properly fix this yet in nbsphinx, because the nbconvert people might still make further breaking changes before nbconvert version 6 is released, see https://github.com/jupyter/nbconvert/issues/1237.

As a temporary measure, I could restrict nbconvert to <6 in the nbsphinx dependencies, but I'm not sure that's the right thing to do. What do you think?

[mgeier]

This has been fixed in nbconvert in the meantime.

[HealthyPear]

I am getting this exact error but from a sphinx build

Sphinx parallel build error:
AttributeError: module 'nbconvert.preprocessors.execute' has no attribute 'CellExecutionError'

I am using

nbconvert                 7.13.0             pyhd8ed1ab_0    conda-forge
nbconvert-core            7.13.0             pyhd8ed1ab_0    conda-forge
nbconvert-pandoc          7.13.0             pyhd8ed1ab_0    conda-forge
nbsphinx                  0.9.3              pyhd8ed1ab_0    conda-forge

both seem to be at the latest release

Let me know if you'd prefer I open a new issue

[HealthyPear]

from a python prompt this works

from nbconvert.preprocessors import CellExecutionError

but this doesn't

from nbconvert.preprocessors.execute import CellExecutionError

as the error indeed reports

[HealthyPear]

see #772

[mgeier]

This seems to have been removed in https://github.com/jupyter/nbconvert/pull/2079, maybe unintentionally.

But since there has been a new nbconvert release since, I guess we should fix it. Thanks for providing the fix in #772!

[mgeier]

A new patch-release of nbconvert has just been released, which should fix this again.

I think we should still change the import on our side.